FR3071110B1 - METHOD FOR MONITORING AND CONTROLLING AN ELECTRICAL NETWORK - Google Patents

Abstract

The invention relates to a method for monitoring and controlling an electrical network which comprises at least one transformation station (P) delimiting a medium-voltage sub-network and a low-voltage sub-network and several entities (Ei) connected to a network. determined voltage in the low voltage subnetwork, each entity being a consumer and / or producer of electricity in the low voltage subnetwork. The method consists in determining control data (Dx) in production and / or in electrical consumption to be applied to one or more entities of the low voltage subnetwork, taking into account previously determined production and / or consumption quotas.

Description

METHOD FOR MONITORING AND CONTROLLING AN ELECTRICAL NETWORK

Technical field of the invention

The present invention relates to a method for monitoring and controlling an electrical network that includes at least one transformer station between a medium voltage subnetwork and a low voltage subnetwork. The invention also relates to the monitoring and control system for implementing said method.

State of the art

There are many solutions today that can monitor and control an electrical network, especially to ensure flexibility. Today, it is in fact to take into account both the power consumption of each entity of the electricity grid, but also the possible electricity production of different network entities (via solar panels for example) and constraints. physical network (overvoltage, undervoltage, overcurrent ...). According to these constraints, it will indeed sometimes be necessary to act on the production and / or consumption of one or more entities of the network.

Traditionally, to achieve such control of an electrical network, a control center is based for example on the electrical grid and uses "load flow" type or "optimal power flow" methods that determine the power flow in the various branches of the power grid to possibly act on the entities of the network.

However, the network plan is not always available or may be wrong, which will inevitably lead to control errors. In addition, previous solutions are not always adapted to all the operating situations encountered. They can thus act on the network in an inappropriate way, even counterproductive. Concrete examples will be presented at the end of the description.

The document entitled "Efficient Computation of Sensitivity Coefficients of Node Voltages and Line Currents in Unbalanced Radical Electrical Distribution Networks" -Konstantina Christakou, Jean-Yves Le Boudee, Mario Paolone, Dan-Cristian Tomozei, describes an optimized principle of network control, based on on the creation of influence matrices for each operating point, from a map of the network.

This prior solution is however unreliable, linked to the fact that it relies on a network plan to build the matrices and difficult to implement because it requires a separate matrix for each operating point.

The object of the invention is therefore to propose a method for monitoring and controlling an electrical network, which is reliable and effective for implementing, at the level of the entities of an electrical network, actions adapted to the constraints of the electrical network and which does not require the use of a layout of the electrical network to operate, thus avoiding an unsuitable operation in the event of an error on a network plan.

Presentation of the invention

This object is achieved by a method for monitoring and controlling an electrical network that includes at least one transformer substation delimiting a medium voltage subnetwork and a low voltage subnetwork and several entities connected to a determined voltage in the sub-network. low-voltage network, each entity being a consumer and / or producer of electricity in the low-voltage sub-network, said method comprising:

A step of acquiring an electric power consumed and / or produced by each entity of the low voltage sub-network at successive time steps,

A determination step at each new time step of a production and / or consumption quota to be allocated to each entity,

A step of determining a minimum value and a maximum value of production and / or consumption quota to be respected for each entity taking into account the electrical power consumed and / or produced at a time step preceding said new step time, a step of determining an electrical power consumed and / or produced forecast for each entity, based on said electrical power consumed and / or produced acquired at a time step preceding said new time step,

Said production and / or consumption quota being a solution to an optimization problem taking into account: At least one unique data model acquired by learning, applied at each new time step and which comprises data representative of a influence of the electrical power consumed and / or produced by each entity on at least one monitored electrical quantity,

Technical constraints to respect for said at least one monitored quantity,

Of said minimum value and maximum value of production and / or consumption quota determined,

Of said consumed and / or projected electrical power determined for each entity, of an optimization function chosen to allocate the production and / or consumption quotas to the entities,

A step of determining control data in production and / or in electrical consumption to be applied to one or more entities of the low voltage subnetwork taking into account each production and / or consumption quota determined.

According to a particular aspect of the invention, said at least one monitored electrical quantity corresponds to the voltage standard of each entity at its point of connection.

According to another particular aspect of the invention, said constraints to be respected for said at least one monitored electrical quantity correspond to a minimum voltage value and a maximum voltage value for each entity.

According to another aspect of the invention, said at least one monitored electrical quantity corresponds to the electrical power passing through the transformer station.

According to another aspect of the invention, said constraints to respect for said at least one monitored electrical quantity correspond to a minimum value of power transiting to the transformer and a maximum value of power transiting at the transformer station (P).

According to another aspect of the invention, said at least one monitored electrical quantity corresponds to the standard of the intensity passing through a cable.

According to another aspect of the invention, said optimization function is chosen from: - Maximizing the sum of the production and / or consumption quotas allocated to the entities, - Maximizing the equity of the consumption restrictions between the entities, or - Maximize the logarithm of the power assigned to each entity (Ei).

According to another aspect of the invention, the step of determining a minimum value and a maximum value of production and / or consumption quota to be respected for each entity is carried out taking into account technical data relating to each entity, selected from one or more of the following data:

A level of power attributed to each entity according to its supply contract,

Data relating to the power generation solution installed in each producing entity,

A maximum amount of energy or cleavable power per entity,

Data relating to the storage capacity of the entity if it is producing.

According to another aspect of the invention, the method comprises a step of determining the number of new electrical data of the electrical network and a step of updating said at least one data model of the electrical network when sufficient electrical data are acquired available. . The invention also relates to a system for monitoring and controlling an electrical network which comprises at least one transformer station delimiting a medium voltage subnetwork and a low voltage subnetwork and several entities connected to a determined voltage in the sub-network. low-voltage network, each entity being a consumer and / or producer of electricity in the low-voltage sub-network, said system comprising:

A module for acquiring electrical power consumed and / or produced by each entity of the low voltage sub-network at successive time steps,

A control module configured to: Determine at each new time step a production and / or consumption quota to be allocated to each entity, Determine a minimum value and a maximum value of production and / or consumption quota to be respected for each entity taking into account the electrical power consumed and / or produced at a time step preceding said new time step, Determine a consumed electric power and / or expected output for each entity, based on said electrical power consumed and / or product acquired at a time step preceding said new time step, - said quota (q) of production and / or consumption being a solution to an optimization problem taking into account: At least one single data model acquired by learning, applied at each new time step and which includes data representative of an influence of the electric power consumed and / or produced by each entity on at least one supervised electrical quantity,

Technical constraints to respect for said at least one monitored quantity,

Of said minimum value and maximum value of production and / or consumption quota determined,

Of said electrical power consumed and / or projected output determined for each entity, of an optimization function chosen to allocate the production and / or consumption quotas to the entities, to determine control data in production and / or electricity consumption at apply to one or more entities of the low voltage subnetwork taking into account each production and / or consumption quota determined.

According to a feature of the system, said at least one monitored electrical quantity corresponds to the voltage standard of each entity at its point of connection.

According to another feature of the system, said constraints to respect for said at least one monitored electrical quantity correspond to a minimum voltage value and a maximum voltage value for each entity.

According to another feature of the system, said at least one monitored electrical quantity corresponds to the electrical power passing through the transformer station.

According to another particularity of the system, said constraints to respect for said at least one monitored electrical quantity correspond to a minimum value of power transiting to the transformer and a maximum value of power transiting at the transformer station.

According to another particularity of the system, said at least one monitored electrical quantity corresponds to the standard of the intensity passing through a cable.

According to another feature of the system, said optimization function is chosen from: - Maximize the sum of the production and / or consumption quotas allocated to the entities), - Maximize the equity of the consumption restrictions between the entities, or - Maximize the logarithm of the power assigned to each entity.

According to another particularity of the system, the control module is configured to determine a minimum value and a maximum value of production and / or consumption quota to be respected for each entity, taking into account technical data relating to each entity, selected from a or more of the following data:

A level of power attributed to each entity according to its supply contract,

Data relating to the power generation solution installed in each producing entity,

A maximum amount of energy or "clippable" power per entity,

Data relating to the storage capacity of the entity if it is producing.

According to another feature of the system, it comprises a learning module configured to determine a number of new electrical data of the electrical network and an update of said at least one data model of the electrical network when sufficient electrical data acquired are available.

BRIEF DESCRIPTION OF THE FIGURES Other features and advantages will appear in the following detailed description with reference to the accompanying drawings in which:

Figure 1 schematically shows a simplified power grid to which the monitoring and control method of the invention can be applied and the monitoring and control system of the invention used to monitor such an electrical network.

Figure 2 schematically shows the monitoring system and control of the invention and its principle of operation.

Figures 3 and 4 show different operating cases to which the method of the invention applies.

Detailed description of at least one embodiment

With reference to FIG. 1, the invention applies to an electrical network that comprises at least one MT / LV type transforming station P delimiting a low-voltage sub-network BT and a medium voltage sub-network MT.

The low-voltage sub-network comprises several entities Ei (of rank i, with i ranging from 1 to n and n greater than or equal to 2) which will each be of consumer type and / or type producing an electric power.

By entity Ei, in a nonlimiting manner, it is necessary to understand, for example, a dwelling, a group of dwellings, at least one factory, at least one power plant, for example of photovoltaic type, wind turbine or a combination of these different entities, for example a house with photovoltaic panels to produce electricity. Each entity Ei will include or be associated with a control unit responsible for implementing appropriate actions taking into account the control data received, for example clipping actions in production and consumption. A clipping action will consist in a limitation in production or consumption of an entity. By way of example, the electrical network represented in FIG. 1 thus comprises n entities E 1.

In a nonlimiting manner, the network represented in the figure comprises n entities Ei connected to the transformation station P via a network branch. It should be understood that the low voltage sub-network may comprise several branches and that each network branch is single-phase or multiphase.

In the low voltage subnetwork, each entity Ei is characterized by the following information:

Producer and / or consumer of an electrical power;

Identity of the connection phase of the entity in the low voltage subnetwork;

FIG. 2 represents a system for monitoring and controlling such an electrical network.

This monitoring and control system comprises a CPU processing and command unit responsible for implementing the following functions:

Execution of a measurement data acquisition module D1;

Running an M1 learning module to build and update an MD data model;

Execution of an M2 control module to determine a suitable control of the electrical network;

Execution of a module for recording various data related to each entity Ei and to the transformation station P of the electrical network;

The modules are software modules configured to be run by the CPU processing and control unit. They are stored on any known and executable computer support. The CPU processing and control unit typically comprises a microprocessor and storage means. It advantageously comprises several inputs intended to be connected to measuring means, such as sensors and outputs intended to be connected to control units each dedicated to the control of a separate entity Ei. Of course, other devices may be connected to the inputs / outputs of the CPU processing and control unit.

The various data recorded by the recording module can be:

D10 technical data attached to the electrical network or,

D20 data relating to each entity of the low voltage subnetwork.

The technical data D10 may be acceptable limit values for the transformation station P and / or acceptable minimum and maximum mains voltage.

The D20 data for each entity can be the power level assigned to each entity according to its supply contract, data relating to the power generation solution installed in each producing entity (for example the surface of solar panels, etc.). ), the maximum amount of cleavable energy or power per entity and / or data relating to the storage capacity of the entity Ei if it is a producer ...

The monitoring and control system may include an HMI human-machine interface for parameterizing the system and in particular the entry for recording data D10, D20 linked to each entity Ei of the low voltage subnetwork.

The monitoring and control system may comprise means for measuring one or more electrical quantities at each entity of the low voltage subnetwork. These measuring means are connected to said data acquisition module.

These measuring means are advantageously current (to determine each current Ii) and voltage (to determine each voltage Vi) properly positioned sensors, in particular on the main supply circuit of each entity Ei and the connection point of each entity Ei on the low voltage subnetwork. Any other sensor can be used, such as energy sensors to calculate the electrical energy consumed by the entity and send it to a dedicated central unit (a data collection server) or directly to the UC processing and control unit. The execution of the control module M2 is configured to implement at least part of the monitoring and control method of the invention.

According to one particular aspect of the invention, the monitoring and control method of the invention thus consists in monitoring the electrical network in order to mainly determine consumption and / or production control data to be sent to one or several entities Ei of the sub-network low voltage. Other control data could be sent such as for example data:

Control for energy storage for a producing entity having a storage capacity that is free,

Control for release of energy to one or more generating entities of the network, if the supply through the network was cut or limited upstream of the transformation station P,

Control for a modification of disjunction threshold in consumption and / or in production.

In order to function, the control module M2 is based on at least one data model MD1 and / or MD2 obtained at the end of a learning process and that can be updated regularly when sufficient new data acquired (for example by means of measures) become available. The training is implemented by the learning module M1 executed by the processing and control unit UC. At the end of the training, each data model MD1, MD2 is unique for a set of operating points.

To create the data model MD1, MD2, the learning module M1 relies on data acquired by the processing and control unit UC.

These data can be:

D2 data from a history of sensor measurements at the transformation station P;

D3 data from a history of sensor measurements at the entities Ei of the network;

D4 data from a history of various sensor measurements, for example on different branches of the network;

The data D1 from the last measurements made by all the sensors of the network at the different branches of the network, at the level of each entity Ei of the network and / or at the level of the transformation station;

In case of absence of data history, the data are acquired by measurement, thanks to the measuring means described above, over a longer or shorter learning period according to the size of the electricity grid to be monitored and according to the period sampling. This duration can for example be 24h, several days or even several months. The measurement acquisition sampling period can be adjustable.

To construct the data model MD1, MD2, the data used by the processing and control unit UC (from a history and / or measured on the learning duration) are at least the following:

The active power consumed and / or produced by each entity Ei of the low voltage subnetwork;

The standard of the voltage of each entity Ei;

The electrical power passing through the transformer station P of the network;

These data are acquired over time, thus defining several successive operating points.

Advantageously, the data that is acquired are synchronized with each other so as to form said operating points. By way of example and in a nonlimiting manner, the data acquired are RMS (Root Mean Square) data. As an example, the first table below represents a history of measured data in the low voltage subnetwork, each client corresponding to a separate entity Ei:

In this table, we can see the power consumed or produced (if negative) by each entity and the voltage at their connection point (phase) at different times, each time thus defining a separate operating point.

For example, the second table below represents a history of measured data at the transformer station:

This second table shows the voltage per phase at the secondary (LV side) of the transformer station, as well as the intensity crossing the transformer station (also measured at the secondary) at the same times as those defined above. This intensity can be negative in case of electric generation of one or more entities of the low voltage subnetwork.

Once enough data have been acquired, the learning module M1 applies a learning algorithm on a pair of data, this pair being chosen according to the technical constraints to be respected. For example, it can be one of the following pairs of data:

Active power consumed and / or produced per entity / voltage standard of each entity. This makes it possible to control the voltage at the connection point of each entity.

Active power consumed or produced by entity / power transiting at the transformer station. This ensures that the power to the transformer is correct.

Active power consumed or produced by each entity / standard of intensity passing through a cable. This ensures that the standard of the intensity passing through this cable is correct.

For the processing of the acquired data, the learning module can implement a learning algorithm in the form of a linear regression. Any other adapted algorithm could be used (ex: neural network, ...).

For all operating points, the learning module M1 constructs a single data model, an operating point corresponding to all the measurements taken at a given moment (corresponding to the date column in the tables above).

This gives the data pair a unique data model in the form of a matrix. Depending on the control performed, only one matrix will be used to control this data pair, regardless of the operating point.

More generally, for the first pair of data mentioned above, the learning module M1 thus generates a data model MD1 which comprises: • Data representative of the influence of the power consumed and / or produced by each entity on the voltage standard of the other entities of the low voltage subnetwork;

For the second pair of data mentioned above, the learning module M1 thus generates a data model MD2 which comprises: • Data representative of the influence of the power consumed and / or produced by each entity on the transiting power through the processing station;

Each data model tells us the influence of the power consumed or produced by the entities on the other parameter of the data pair.

In the data model MD1, the matrix (designated Au) thus comprises several columns, each column x being dedicated to an entity Ei distinct from the low voltage subnetwork, and several lines, each line being also dedicated to an entity Ei distinct from the low voltage subnetwork. The columns and rows of the matrix are supplemented by coefficients. In each column, the coefficients Cxy represent the influence of the power consumed or produced by an entity of the low voltage subnetwork referenced in a column on the monitored value of each of the other entities of the low voltage subnetwork identified on each line. By way of example, such a matrix, represented in the form of a table, is the following:

For example, in the case of a linear learning, a column x of the matrix Au represents the influence of the net consumption of an entity Ex of the low voltage subnet on the norm of the voltage of all the other entities of the same network presented on each line. Thus, a coefficient -0.5 in position (x; y) in this matrix indicates that if the Ex entity increases its consumption by 1 kW, the voltage of another entity Ey of the subnet decreases by 0.5 V.

Thanks to the data model, it is thus possible to ensure that the increase of a production (or consumption) quota of one entity is not likely to push the voltage of another entity out of the authorized terminals.

The same operation applies to the MD2 data model (represented by an Ap matrix) in which each column represents the influence of the production or consumption of an entity of the low voltage subnetwork on the power that passes to the transformer station. P.

Of course, other data than those defined above could also be acquired (by measure or otherwise) to enrich the data model. Without limitation, it may be intensity measurements in a particular line of the electrical network, measurements attached to another transformer station included in the electrical network.

During the implementation of the monitoring and control method, the control module M2 is configured to determine the control data to be sent to the entities Ei from the last data D1 acquired, received as input, and data whose UC processing and control unit already has.

For this, the control module M2 determines the qprodi production quota and / or the consumption quota qconsi to allocate to each entity of the electricity network. It sets up a principle of optimization. The optimization principle makes it possible to determine the best quotas to send to the entities Ei of the network. In other words, it will be a question of finding a solution that can satisfy the targeted optimization principle and therefore determine all the quotas that make it possible to fulfill an optimization objective. This objective is defined by an objective function f. For example, it will be necessary to determine the qprodi production quota or the consumption quota q to be allocated to each Ei entity that makes it possible to: - Maximize the sum of the quotas allocated to the entities, - Maximize the fairness of the consumption restrictions between entities, or - Maximize the sum of the logarithm of the power assigned to each entity.

Depending on the one or more pairs of data monitored, the control module M2 relies on one or more of the following measured values over time:

Power (active, possibly reactive) power consumed and / or generated (whether the two are measured independently or only the sum is actually reported) by one or more entities of the network;

Electrical power passing through a branch of the power grid, in an entity Ei and / or at the substation P; - Standard of the voltage Vi of one or more entities of the sub-network low voltage at its point of connection; - Detailed voltage (real part + imaginary part) for one or more entities at their point of connection; - Standard of the intensity flowing in one or more branches of the network;

Detailed intensity (real part + imaginary part) in one or more branches of the network;

The control module M2 operates in successive time steps, preferably all of identical duration. For each new time step, the control module M2 is responsible for determining the qprodi production quota and the consumption quota qconsi to allocate to each entity Ei. The control module M2 determines separately the production quota and the consumption quota to be allocated to each entity. Of course, if no supervised network entity is a producer, determining production quotas will be useless and vice versa.

To determine the qprodi production quota to be allocated to each entity Ei, the control module M2 implements the following steps:

It acquires measurement data D1 of the power consumed and / or produced for each entity Ei of the network; These data are measured at successive time steps (for example every 10 minutes) by the measuring means;

It determines, for each entity Ei, the minimum value qmin (t) and the maximum value qmax (t) of production quota that it can assign to the entity Ei; To set these limit values, it can in particular be based on the production data measured at the previous time step for this entity Ei;

It determines for each entity Ei a forecast of the power consumed pc ^ 7 (t) by each entity Ei for this new time step; This forecast is for example carried out taking into account the power data measured at one or more previous time steps;

He inserts these data into an optimization problem taking into account the different constraints set in the previous steps;

Solving the optimization problem allows it to obtain the qprodi production quota to allocate to each entity Ei of the network;

Optionally, the quota obtained for each entity Ei can be corrected;

These different steps are implemented identically to determine the consumption quota qconsi to allocate to each entity Ei, the power consumption forecast pc ^ (t) being replaced by a prediction of the power produced pp ^ 7 (t) .

In the optimization problem, the control module relies on at least one data model (MD1 or MD2) learned during the learning phase, corresponding to the pair of data monitored.

Below, we describe the optimization problem managed by the M2 control module when determining production quotas and consumption quotas to allocate n entities Ei network. Without limitation, these optimization problems are established on the basis of the two data models MD1 and MD2 taken in combination, but it should be understood that they could be established on the basis of one of the two models only if a only couple of data was monitored. Conversely, other data models corresponding to other monitored data pairs could also be integrated. The reasoning is carried out for the n entities Ei. - Production quota

For production, the optimization problem to be solved is as follows:

(1)

With: - q the production quota to be allocated for the time step t for the n network entities (q with several qprOd_i quotas); - f the objective function which depends on the quota allocated to each entity; - Pcons (i_l) and Pprod (t_l) which are the power measurements in consumption and in production at the previous time step (t - 1); - Pcôns (F) which is a forecast of the consumption of the entities of the network for the time step to come (t). This prediction may depend on other measured values (in particular pC £ ms (t - 1) and pprOd (t ~ 1));

umm and umax the minimum value and the maximum value of acceptable voltage for each entity of the network, corresponding to data D10 mentioned above; - Ptrans and Ptrans the minimum value and the maximum acceptable power value at the transformer, corresponding to data D10 mentioned above; - qmin (t) and qmax (t) which correspond to the minimum value and the maximum quota value attributable to each entity. They depend on time and are adjusted before solving equation (1) above. They can be chosen according to the measurements made at the previous time step (pcons (t - 1) and pprOci (t ~ 1))> but also various information (data D20) on the entities and their electrical installations (peak power of solar installations, power acceptable by their circuit breaker ...) or contractual data. C4p, hp) and (4u, hu) which are the matrices of the data models MD1, MD2 learned during the learning phase. - Consumption quota

For consumption, the optimization problem to be solved is as follows:

(2)

With: - q the consumption quota to be allocated for the time step t for the n entities of the network (q with several qCOnsj quotas)! - f the objective function which depends on the quota that is allocated to each entity. It can be different from the function f of equation (1);

- Ppràd (t) which is a forecast of the consumption of the entities of the network for the time step to come (t). This prediction may depend on other measured values (in particular pcons (t - 1) and pprOd (t ~ 1)); umin and umax the minimum value and the maximum value of acceptable voltage for each entity of the network, corresponding to data D10 mentioned above; - Ptrans and Ptrans the minimum value and the maximum acceptable power value at the transformer, corresponding to data D10 mentioned above; - qmin (t ') and qmax (t) which correspond to the minimum value and the maximum value of allowances attributable to each entity. They are different from those defined for production but depend on the same factors. They can indeed be chosen according to the measurements made at the previous time step (pcons (t - 1) and pprOd (t ~ 1)), but also various information (data D20) on the entities and their electrical installations (power peak solar installations, power acceptable by their circuit breaker ...) or contractual data. (Ap, bp) and (Au, hu) which are the matrices of data models MD1, MD2 learned during the learning phase.

In a non-limiting way, the "objective" function consists, for example, in optimizing electrotechnical criteria such as reducing losses, maximizing the sum of the power produced by the entities, maximizing the minimum quota allocated to the entities Ei, maximizing the logarithm the power granted to each entity ...

The control module M2 then determines the control data Dx to be sent to one or more of the entities according to the quota in production and / or consumption (qprodi, qCOns_i) determined for each entity of the network thanks to the reasoning described herein. -above. From the control data received, a control unit associated with each entity Ei will be able to carry out appropriate actions (clipping, shedding ...).

With reference to FIGS. 3 and 4, several examples of controls implemented are detailed below.

First case of operation - Figure 3

Consider the simplified network of Figure 3. It consists of at least the four entities E1-E4 drawn, connected on a single branch of the network. The first three entities are consumers and producers and the E4 entity is only a consumer. The entity E1 is connected to the phase (a) of the low voltage sub-network; The entity E2 is connected to the phase (b) of the low voltage sub-network; The entity E3 is connected to the phase (c) of the low voltage sub-network; The entity E4 is connected to the phase (a) of the low voltage sub-network;

On a day of great sun, the solar panels of the entities E1, E2, E3 inject a lot of power on the network and consequently the tension along the main line increases to reach its maximum on the phase (a) at the level of the entity E4.

In this case of operation, the data model MD1, represented in the form of the matrix Au, would be the following:

A common strategy would be to clip the nearest producing entity to the E4 entity, that is, the E3 entity. But the entity E3 is connected to phase (c). A capping of its production would therefore lead to a (slight) increase in the voltage on phase (a) instead of decreasing it.

It can be seen that on the fourth column of the matrix Au, the increase of one watt in the consumption of the entity E4:

• reduces the voltage of the entity E1 by 0.2 V, • increases the voltage of the entity E2 by 0.04 V, • increases the voltage of the entity E3 by 0.03 V, • reduces the voltage of the 0.4V E4 entity

To reduce the voltage of the entity E4, it is therefore necessary to act preferentially on the production of the entity E4 (0.4> 0.3). But as the entity E4 has no means of production, it is therefore necessary to act preferably the entity E1 to effectively reduce the voltage of the entity E4.

Thanks to the MD data model, the monitoring and control method of the invention is able to take into account the increasing or decreasing influences of one entity over another and is therefore configured to determine on which entity, an action of clipping must be carried out.

Second Case of Operation - Figures 4A, 4B and 4C

Consider the simplified network of Figure 4A in which: E1 to E4 = Consumer E5 and E6 = Consumer + Producer

The low voltage subnetwork comprises a main branch 1 on which are connected the entities E1 and E2 and two parallel secondary branches 10, 11 connected at a point of connection to the main branch, a first secondary branch 10 on which the entities are connected. E3 and E4 and a second secondary branch 11 on which the entities E5 and E6 are connected.

For simplicity, all entities E1 to E6 are connected to the same phase of the low voltage subnetwork.

In strong sunlight, entities E5 and E6 produce a lot of electrical energy, which they send to the power grid. This has the consequence of increasing the voltage on the second secondary branch 11. On the other hand, if the entities E3 and E4 consume a lot, the voltage on the first secondary branch 10 tends to decrease. This is illustrated by the profile PO shown in FIGS. 4B and 4C. On this profile, it can be seen that, in the situation mentioned above, the entity E6 has a voltage at its connection point which is greater than a voltage limit (VH) and the entity E4 has a voltage at its connection point that is less than a voltage limit (VL).

If both effects occur simultaneously, the voltage can no longer be simply adjusted with a tap-change transformer as usual.

Figure 4B illustrates the answer that could be given today with a tap-change transformation. A transformer with a change of crisis would either solve under-voltage problems by aggravating the problems of overvoltage (profile P1 dotted-entity E4 returns above VL - entity Ξ5 exceeds the limit VH) or to adjust overvoltage problems aggravating the undervoltage problems (profile P2 in gray line-entity E6 returns below the limit VH -entity E3 goes below the limit VL).

In this example, the Au matrix could look like this:

From this matrix, it is possible to note that the two entities E5 and Ξ6 have very little influence on the two entities E3 and E4 and vice versa, this thanks to the weakness of the coefficients present in the matrix. On the basis of these coefficients, solving the optimization problem will not choose to clip the output of the entity Ξ3 or E4 in case of overvoltage at the entities E5 or E6 (and vice versa).

The solution of the invention also makes it possible to learn that the production of the entities E5 and E6 causes an increase in the voltage on their branch while the consumption of the entities E3 and E4 cause a decrease in the voltage on their channel. By determining on which entities the clipping and erasure actions must be carried out, the solution of the invention makes it possible to obtain a voltage profile (Profile D3-FIG. 4C) making it possible to fit all the entities Ei within the limits of operation of the low voltage sub-network.

It is understood that the solution of the invention thus has many advantages, among which:

It does not require any reconstruction of the electrical network plan, as is often done in the state of the art; 5 - It can allow control of the electrical network from a single data model, adapted to the pair of data monitored; 3

Claims (7)

1. A method of monitoring and controlling an electrical network which comprises at least one transformer station (P) delimiting a medium voltage subnetwork and a low voltage subnetwork and several entities (Ei) connected to a determined voltage in the low-voltage sub-network, each entity being a consumer and / or producer of electricity in the low-voltage sub-network, characterized in that said method comprises: a step of acquisition of an electric power consumed and / or produced by each entity (Ei) of the low voltage subnetwork at successive time steps, A step of determining at each new time step a production and / or consumption quota (q) to be allocated to each entity (Ei) , A step of determining a minimum value (qmin (t)) and a maximum value (qmax (t)) of production and / or consumption quota to be respected for each entity (Ei) taking into account the electrical power con summed and / or produced at a previous time step (pCOns (t-1)> Pprodtt-1)) said new time step, A step of determining an electric power consumed and / or produced forecast (pc ^ (t ), Pp ^ Δfty) for each entity, based on said consumed and / or produced electrical power acquired at a time step preceding said new time step, Said quota (q) of production and / or consumption being a solution to a optimization problem taking into account: At least one data model (MD1, MD2) acquired by learning, applied to each new time step and which includes data representative of an influence of the electrical power consumed and / or produced by each entity (Ei) on at least one monitored electrical quantity, technical constraints to respect for said at least one monitored quantity, of said minimum value and maximum value of production and / or consumption quota determined , Of said electrical power consumed and / or projected output determined for each entity, - Of an optimization function chosen to allocate the production and / or consumption quotas to the entities, - A step of determination of control data (Dx) in production and / or in electricity consumption to be applied to one or more entities of the low voltage subnetwork taking into account each production and / or consumption quota determined. 2. Method according to claim 1, characterized in that said at least one monitored electrical quantity corresponds to the voltage standard of each entity at its point of connection. 3. Method according to claim 2, characterized in that said constraints to be observed for said at least one monitored electrical quantity correspond to a minimum value (un, in) of voltage and a maximum value (u ', iax) of voltage for each entity (Ei). 4. Method according to claim 1, characterized in that said at least one monitored electrical quantity corresponds to the electrical power passing through the transformer station. 5. Method according to claim 4, characterized in that said constraints to respect for said at least one monitored electrical quantity correspond to a minimum value (ptraks) of power transiting to the transformer and a maximum value (p ai ai) of power transiting at level of the transformer station (P). 6. Method according to claim 1, characterized in that said at least one monitored electrical quantity corresponds to the standard of the intensity passing through a cable. 7. Method according to one of claims 1 to 6, characterized in that said optimization function is chosen from: - Maximize the sum of the production and / or consumption quotas allocated to the entities (Ei), - Maximize the equity of consumption restrictions between entities, or - Maximize the logarithm of the power attributed to each entity (Ei). 8. Method according to one of claims 1 to 7, characterized in that the step of determining a minimum value (qmiîl (t)) and a maximum value (qmax (t)) production quota and / or of consumption to be respected for each entity (Ei) is carried out taking into account data (D20) techniques relating to each entity (Ei), chosen from one or more of the following data: A power level attributed to each entity (Ei) ) according to its supply contract, data relating to the power generation solution installed in each producing entity, a maximum amount of energy or cleavable power per entity, data relating to the storage capacity of the entity (Ei ) if it is a producer. 9. Method according to one of claims 1 to 8, characterized in that it comprises a step of determining the number of new electrical data of the electrical network and a step of updating said at least one data model (MD1, MD2) when sufficient electrical data is available. 10. System for monitoring and controlling an electrical network that includes at least one transformer station (P) delimiting a medium voltage subnetwork and a low voltage subnetwork and several entities (Ei) connected to a determined voltage in the low voltage sub-network, each entity being a consumer and / or producer of electricity in the low voltage sub-network, characterized in that said system comprises: A module for acquiring an electric power consumed and / or produced by each entity (Ei) of the low voltage subnetwork at successive time steps, A control module (M2) configured to: Determine at each new time step a quota (q) of production and / or consumption to be allocated to each entity (Ei), Determine a minimum value (qmiîl (t)) θί a maximum value (qmax (t)) of production and / or consumption quota to be respected for each entity (Ei) taking into account the electrical power ique consumed and / or produced at a step of previous time (pccms (t-1), Pprodit ~ 1)) said new time step, Determine a consumed electric power and / or produced forecast (pc ^ (t), Pp ^ â (t)) for each entity, based on said electrical power consumed and / or produced acquired at a time step preceding said new time step, - said quota (q) of production and / or consumption being a solution to a optimization problem taking into account: At least one data model (MD1, MD2) acquired by learning, applied to each new time step and which includes data representative of an influence of the electrical power consumed and / or produced by each entity (Ei) on at least one monitored electrical quantity, Technical constraints to respect for said at least one monitored quantity, of said minimum value and maximum value of production and / or consumption quota determined, De ladi the electrical power consumed and / or projected output determined for each entity, an optimization function chosen to allocate production and / or consumption quotas to entities, determine control data (Dx) in production and / or in consumption to apply to one or more entities of the low voltage subnetwork taking into account each production and / or consumption quota determined. 11. System according to claim 10, characterized in that said at least one monitored electrical quantity corresponds to the voltage standard of each entity at its point of connection. 12. System according to claim 11, characterized in that said constraints to respect for said at least one monitored electrical quantity correspond to a minimum value (umin) of voltage and a maximum value (umax) of voltage for each entity (Ei). 13. System according to claim 10, characterized in that said at least one monitored electrical quantity corresponds to the electrical power transiting at the transformer station. 14. The system of claim 13, characterized in that said constraints to be observed for said at least one monitored electrical quantity correspond to a minimum value (PtÏÏSis) of power transiting to the transformer and a maximum value (p ^ years) of power transiting at level of the transformer station (P). 15. System according to claim 10, characterized in that said at least one monitored electrical quantity corresponds to the standard of the intensity passing through a cable. 16. System according to one of claims 10 to 15, characterized in that said optimization function is chosen from: - Maximize the sum of the production and / or consumption quotas allocated to the entities (Ei), - Maximize the equity of consumption restrictions between entities, or - Maximize the logarithm of the power attributed to each entity (Ei). 17. System according to one of claims 10 to 16, characterized in that the control module (M2) is configured to determine a minimum value (qmin (t)) and a maximum value (qmax (t)) of the quota of production and / or consumption to be respected for each entity (Ei) taking into account data (D20) relating to each entity (Ei), chosen from one or more of the following data: A power level assigned to each entity (Ei) ) according to its supply contract, data relating to the power generation solution installed in each generating entity, maximum amount of energy or "clippable" power per entity, data relating to the storage capacity of the entity (Ei) if she is a producer. 18. System according to one of claims 10 to 17, characterized in that it comprises a learning module (M1) configured to determine a number of new electrical data of the electrical network and an update of said at least one model data (MD1, MD2) from the power grid when sufficient electrical data is available.